TUSHAR JUNEJA

Workflow Orchestration Engine

• Built and operated an event-driven workflow engine processing millions of events/day across multiple microservices, powering financial data pipelines for banks, NBFCs, and fintechs in India's Account Aggregator ecosystem.
• Eliminated head-of-line blocking and consumer group collapse by implementing partition-level pause/resume, sequential offset tracking, and per-run thread affinity across Kafka consumers.
• Introduced topic-level workload segregation (BATCH / EXPRESS / STANDARD), achieving isolated SLAs and zero cross-workload interference in a multi-tenant pipeline.

Tech: Java · Spring Boot · Kafka (1,024 partitions, MSK) · PostgreSQL · Debezium CDC · Redis · EKS

SMS Analytics Service

• Built an end-to-end SMS analytics pipeline converting unstructured Indian bank SMS into structured financial signals — powering credit decisioning, PFM insights, and early warning systems for lenders and insurers.
• Designed a three-layer processing engine (real-time extraction → classification → batch aggregation) with a self-improving template discovery pipeline, reducing manual rule maintenance.

Tech: Python · DuckDB · PostgreSQL · Redis · Kafka (MSK) · S3 Parquet · EKS

Mutual Fund & Equity Data Service

• Built a low-latency FastAPI service serving historical OHLC and NAV data for equities and mutual funds with sub-50ms P95 response times via multi-tier caching (Redis → PostgreSQL) and fully async I/O, minimizing external API dependency.

Tech: Python · FastAPI · PostgreSQL · Redis · asyncio/aiohttp · APScheduler

Workflow & Integration Engine

• Architected an event-driven workflow integration engine for third-party applications, processing thousands of daily events across services — enabling seamless automation between customer platforms and internal systems.

Tech: Node.js · PostgreSQL · Redis · Apache Kafka

AI-Powered Feedback Analytics

• Built an intelligent search and analytics system for a customer feedback platform, enabling product teams to discover and act on user insights across multiple channels.
• Designed a semantic search engine using BERT/Transformers and Azure AI Search, enabling natural language querying and contextual filtering across unstructured feedback data.
• Shipped a GPT-4 powered auto-categorization pipeline with a self-improving feedback loop driven by continuous customer input — reducing manual triage effort for product teams.

Tech:Python · Flask · Celery · BERT/Transformers · GPT-4 · Azure AI Search · Azure Language Services

Infrastructure & Platform

• Built scalable infrastructure supporting secure authentication, custom domain provisioning, and automated login flows.

Tech: AWS (Lambda, CloudFront, ACM) · Cloudflare DNS · Nginx

Leadership & Cross-Functional

• Mentored an 8-member engineering team, driving skill development and product delivery cadence.
• Partnered with cofounders to align technical architecture with company vision and market requirements, while managing high-priority customer escalations to inform the product roadmap.

Settlement Batch Processing

• Developed and maintained backend services for daily transaction settlement workflows, ensuring accurate and timely end-of-day batch processing in high-volume payment environments.

Tech : Java · Spring Boot · Spring Batch · Oracle

• Led end-to-end transformation of a monolithic J2EE application into a modular Spring Boot architecture, owning delivery across design, implementation, and rollout.
• Designed a pluggable caching framework supporting OrientDB and Apache Ignite as interchangeable data stores.

Tech: Java · Spring Boot · Spring MVC · AspectJ · Apache Ignite · OrientDB · JMS · ActiveMQ · IBM MQ

• Developed Oracle-based analytics and MongoDB-powered search functionality, improving data accessibility.
• Contributed to implementation and delivery of HRMS module, enhancing product functionality.
• Technologies: Java, Oracle, MongoDB, REST APIs.